356 BOTANICAL GAZETTE [DECEMBER 
The same fluctuation in the water-supplying power of the soil, 
as shown by the Livingston-Koketsu soil points (fig. 2), is even 
more marked. Both the individual determinations and their ten- 
day means show fluctuations closely parallel to those of the growth 
water percentage data. The close “bunching” of the ten-day 
means during July and the period ending September 8 is especially 
suggestive. Since the soil point method was devised in an effort 
to determine directly the water-supplying power of any kind of soil, 
independently of its wilting coefficient or any other physical con- 
stant, this shows clearly that during such periods of drought all the 
soils in the locations studied, save one, dropped to a very critical 
water-delivering power. In the Livingston-Koketsu experiments 
with wheat and Coleus under winter greenhouse conditions, perma- 
nent wilting ensued when the water-delivering power of a soil had 
fallen to a point between 0.04 and 0.11 gm. per two-hour period (as 
compared with about 15.0 gm. for the same period in a nearly satu- 
rated soil). The soil points used in the present determinations, 
however, have a somewhat greater absorbing power. According to 
data as yet unpublished, obtained by H. C. Dreut, the new points 
have 1.25 times the absorptive power of those used by LIvINGSTON 
and Koxetsu. The water-delivering power of the soil at the 
permanent wilting point should therefore be between 0.05 and 
0.14 gm. for a two-hour period. This is the case, at least, if we do 
not take into account the evaporating power of the air at permanent 
wilting. Since, however, the evaporating power of the air through- 
out the droughty periods was greater at the stations considered than 
it was in Lrvincston’s greenhouses, we are safe in doing so, and in 
taking as an approximate water-supplying power at wilting point 
0.15 gm. for the two-hour period. It will be seen that during the 
drought periods all of the stations save one either approached or 
passed this critical point, and that several of them were well beneath 
it for a period of thirty days during July and one of twenty during 
August, with an interval of only ten days between these two pro- 
longed droughts. Further, a comparison with figs. 3 and 5 
show that these were the periods of greatest stress from extremes of 
temperature and evaporation. It is fairly evident, therefore, that 
seedlings that are to survive throughout most of the park must be 
